logLik.tsmarch: Extract Log-Likelihood

In tsmarch: Multivariate ARCH Models

Extract Log-Likelihood

Description

Extract the log likelihood of the model at the estimated optimal parameter values.

Usage

## S3 method for class 'cgarch.estimate'
logLik(object, ...)

## S3 method for class 'dcc.estimate'
logLik(object, ...)

## S3 method for class 'gogarch.estimate'
logLik(object, ...)

Arguments

object	an estimated object from one of the models in the package.
...	none

Details

For all models in the package, the log-likelihood is a combination of the univariate log-likelihood and the multivariate log-likelihood. For the GOGARCH model, being an independent factor model, this is the sum of the univariate GARCH log-likelihoods plus a term for the mixing matrix. The number of parameters in the GOGARCH model reported (“df”) represents the univariate independent factor parameters plus the number of parameters in the rotation matrix U of the ICA algorithm.

Value

An object of class “logLik” with attributes for “nobs” and “df”. The latter is equal to the number of estimated parameters, whilst the former is the number of timesteps (i.e. the number of observations per series).

Author(s)

Alexios Galanos

tsmarch documentation built on April 3, 2025, 7:40 p.m.